Describable Visual Attributes for Face Verification and Image Search
|
|
- Polly Blair
- 6 years ago
- Views:
Transcription
1 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 1 Describable Visual Attributes for Face Verification and Image Search Kumar, Berg, Belhumeur, Nayar. PAMI, Ryan Lei 2011/05/05 Course ammai
2 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 2 Outline Title explanation Introduction Related work Creating labeled image datasets Learning visual attributes Application to face verification Application to face search Possible improvements
3 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 3 Title Explanation Describable visual attributes: Labels that can be given to an image to describe its appearance. Examples: gender, age, skin color, hair style, smiling. Goal 1 - Face verification (authentication): a. Determine whether two faces are of the same individual. b. Knowing the identity, judge if the query face image is the same person. Goal 2 - Image Search: Attribute-based image search system. Sample query: smiling Asian man with glasses.
4 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 4 Introduction (1) Traditional face recognition research: Use of low-level image features to directly train classifier for the classification task. Examples: color (histogram, GCM), texture (Gabor), intensity (Haar, LBP), gradient (SIFT, HOG). The feature representations are often high dimensional, and are in an abstract space. Proposal - attribute-based representation: a. Describable visual attributes ( attributes in this slide) b. Similes: similarities to reference faces.
5 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 5 Recap: semantic concept detection (course MMAI).
6 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 6 Concept space for faces -- (a): attributes
7 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 7 Concept space for faces -- (b): similes
8 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 8 Introduction (2) Advantages of attribute-based representation: A mid-level representation bridging the semantic gap. Dimensionality reduction & manifold discovery. Flexible, generalizable, efficient. Flexible: various levels of specificity. white male a set of people. + brown-hair green-eyes scar-on-forehead a specific person. + smiling lit-from-above seen-from-left a particular image of that person.
9 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 9 Generalizable: Introduction (3) Learn a set of attributes from a large image collection, and then apply them in arbitrary combinations to the recognition of unseen images. Efficient: k binary attributes can identify up to 2 k categories. Requires a much smaller labeled dataset to achieve comparable performance on recognition tasks.
10 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 10 Introduction (4) Contributions of this paper: 1. Introduce attribute & simile classifiers, and face representation using these results. 2. Application to face verification and face search. 3. Releases two large public datasets: FaceTracer and PubFig. Previous work: N. Kumar, et al., Attribute and simile classifiers for face verification, CVPR N. Kumar, et al., FaceTracer: A Search Engine for Large Collections of Images with Faces, ECCV 2008.
11 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 11 Creating Labeled Datasets (1) Steps 1 & 2: Download from a variety of online resources, such as Yahoo! Images and Flickr. Using a commercial face detector (Omron OKAO), detect faces, pose angles, and fiducial points. Using the yaw angle, flip the face so it always faces left. Align faces using linear least square regression.
12 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 12 Creating Labeled Datasets (1) Steps 1 & 2: pose angles fiducial points Download from a variety of online resources, such as Yahoo! Images and Flickr. Using a commercial face detector (Omron OKAO), detect faces, pose angles, and fiducial points. Using the yaw angle, flip the face so it always faces left. Align faces using linear least square regression.
13 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 13 Review of linear least square regression Basics (1-dimensional-output case): Given a vector x, when want to predict its output y. Prepare a training set of N data: {x 1, x 2,..., x N }. Introduce M basis functions Φ j (.), e.g., Gaussian, such that y is a linear combination of M basis components: y = M " w!! (x) = w T!(x) j j j=1 Goal: learn the weight vector w. Prediction: y new = w T!(x new )
14 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 14 Review of linear least square regression Formulation: N Overdetermined systems:! # # X = # data # " basis M " j=1! 1 (x 1 )! 2 (x 1 )!! M (x 1 )! 1 (x 2 )! 2 (x 2 )!! M (x 2 )!! "!! 1 (x N )! 2 (x N )!! M (x N ) Solve the normal equations for w :! j (x i )! w j = y i, i =1, 2,..., N $! & # & # &, y = # & # % " y 1 y 2! y N X T Xw = X T y! w = (X T X) "1 X T y $ & & & & % ' Xw = y
15 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 15 Review of linear least square regression Extension to D-dimensional-output case:! # # W = # # " output dim. w 11 w 12! w 1D w 21 w 22! w 2D!! "! w M1 w M 2! w MD $! & # & # &, Y = # & # % " output dim. y 11 y 12! y 1D y 21 y 22! y 2D!! "! y N1 y N 2! y ND X T XW = X T Y! W = (X T X) "1 X T Y Given 6 fiducial points, predict 6 new coordinates. In our case, W matrix is an affine transformation. $ & & & & %
16 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 16 Creating Labeled Datasets (2) Step 3: Crowd-sourcing part. They chose Amazon Mechanical Turk, a service that matches workers to online jobs created by requesters. Quality control by requiring confirmation of results by several workers, or minimum worker experience, etc. After verification, they collected 145,000 attribute labels, at the cost of USD$6,000.
17 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 17 Amazon Mechanical Turk Worker Requester
18 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 18 Creating Labeled Datasets (3) FaceTracer attribute labels: 15,000 faces with 5,000 labels. (0.33 labels per image) Researchers still need to train their own attribute classifiers to transform the faces into attribute space. PubFig identity labels: 58,797 images of 200 public figures. Development set: 60 people. (for training simile classifiers) Evaluation set: 140 people. (same purpose as LFW) What if jointly labeling identities and attributes?
19 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 19 # of faces of this person.
20 Multimedia Analysis and Indexing, Fall 2010, NTU 20 Attributes labeled in the FaceTracer dataset However, all attribute classifiers in this work are binary.
21 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 21 Learning Visual Attributes (1) Definitions: Consider the attribute gender and the qualitative labels male and female. An Attribute can be thought of as a function that maps an image I to a real value r i. female male r i Large positive (negative) values of r i indicate the present (absent) of the i-th attribute. Core part in this work: Feature selection, though still an open problem in machine learning.
22 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 22 Learning Visual Attributes (2) Low-level features options: a) Region of face. [10] b) Type of pixel value. [5] c) Normalization to apply. [3] d) Level of aggregation to use. [3] A total of 10*5*3*3 = 450 combinations.
23 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 23 Learning Visual Attributes (3) Complete feature pool: However, not all possible combinations are valid, e.g., normalization of hues. a) ˆx = x µ ˆx = x! µ! b) c) d) (concat)
24 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 24 Learning Visual Attributes (4) Forward feature selection in each iteration: 1. Train all possible classifiers by concatenating one feature option with the current feature set. 2. Evaluate the performances by cross-validation. 3. Feature(s) with highest CV accuracy added. 4. Drop the lowest-scoring 70% of features. 5. Keep adding until the accuracy stops improving.
25 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 25 Learning Visual Attributes (5) Classifiers used: SVMs with RBF kernel. (tool: libsvm) Seems to be only one final classifier. ECCV 2008: Boosting approach. Thought: Features A and B work well individually. Will they work well together?
26 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 26 Typically range from 80% to 90%. Analysis shows that all regions and feature types are useful: the power of feature selection.
27 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 27 Learning Visual Attributes (6) Simile classifiers:
28 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 28 Learning Visual Attributes (6) Simile classifiers: 60 reference people. 8 manually selected face regions. (Similes are component-based by definition.) 6 manually selected combinations of [pixel value type, normalization, aggregation], without automatic selection. (Could have done so.) Yield a total of 60*8*6 = 2,880 simile classifiers, each being an SVM with RBF kernel.
29 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 29 Now we have these attributes and similes as midlevel representation. How do we apply them in high-level classification / retrieval problems?
30 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 30 Application to Face Verification (1) Problem: Are these two faces of the same person? Existing methods: Very early work & early work: Compare L2 distance in PCA-reduced space. Improved by the supervised LDA. Early work used well-known low-level features directly. Often make avoidable mistakes: men being confused for women, young people for old, Asians for Caucasians. From this observation, they claim that the attribute and simile classifiers can avoid such mistakes.
31 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 31 Application to Face Verification (2) Many steps have been explained in previous sections. Goal here: the verification classifier.
32 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 32 Application to Face Verification (3) Verification classifier: Train a verification classifier V that compares attribute vectors C(I 1 ) and C(I 2 ) of two face images I 1 and I 2, and returns the decision v(i 1, I 2 ). Define a i = C i (I 1 ) and b i = C i (I 2 ), subscript i meaning the i-th attribute (simile) value. i = 1, 2,, n. Use both the absolute difference and product. 1) a i b i : Observation that a i and b i should be close if the are of the same individual. 2) a i * b i : Observation that a i and b i should have the same sign if the are of the same individual.
33 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 33 Application to Face Verification (4) Verification classifier (cont.): Concatenate them in tuple p i, then concatenate for all i: <p 1,,p n > is the training input to the classifier. Again, they use SVM with RBF kernel. Experiment: Performed on datasets LFW and PubFig. Even the individuals are disjoint in training / testing sets. In other words, machine never sees the same pair of people in its model. Rely solely on attributes / similes.
34 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 34
35 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 35 How well can attributes potentially do? 91.86% 81.57% Attributes obtained by machine learning: 81.57%. Attributes obtained by human labeling: 91.86%. Face verification entirely by human: 99.20%.
36 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 36 Application to Face Search FaceTracer search engine: For details, refer to the work in ECCV Demo video on YouTube: Text-based query. Remember that we have bridged the semantic gap by those mid-level classifiers. For each attribute, search results are ranked by confidences. Convert confidences into probabilities in [0,1] by Gaussian. For multiple query terms, just combine them by taking the product (AND) of those probabilities. Indexing: basic inverted index? Not mentioned.
37 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 37 What about the user gap?
38 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 38 To accurately capture user s search intention: sketch-based concept-based people-attribute -based
39 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 39 Possible Improvements Go beyond binary classifiers for attributes. Regressors are good for quantitative attributes like age. Use feature selection in training the Simile classifiers. Automatic (Dynamic) selection of attributes. Image search: Product of probabilities can t prevent outlier scores. TF-IDF approach for ranking the word frequencies? Combine identity information? Combine location and size, even sketch?
40 Advanced Topics in Multimedia Analysis and Indexing, Spring 2011, NTU. 40 Thank You! Q & A
CS 1674: Intro to Computer Vision. Attributes. Prof. Adriana Kovashka University of Pittsburgh November 2, 2016
CS 1674: Intro to Computer Vision Attributes Prof. Adriana Kovashka University of Pittsburgh November 2, 2016 Plan for today What are attributes and why are they useful? (paper 1) Attributes for zero-shot
More informationShifting from Naming to Describing: Semantic Attribute Models. Rogerio Feris, June 2014
Shifting from Naming to Describing: Semantic Attribute Models Rogerio Feris, June 2014 Recap Large-Scale Semantic Modeling Feature Coding and Pooling Low-Level Feature Extraction Training Data Slide credit:
More informationInteractive Image Search with Attributes
Interactive Image Search with Attributes Adriana Kovashka Department of Computer Science January 13, 2015 Joint work with Kristen Grauman and Devi Parikh We Need Search to Access Visual Data 144,000 hours
More informationDiscriminative classifiers for image recognition
Discriminative classifiers for image recognition May 26 th, 2015 Yong Jae Lee UC Davis Outline Last time: window-based generic object detection basic pipeline face detection with boosting as case study
More informationAn Associate-Predict Model for Face Recognition FIPA Seminar WS 2011/2012
An Associate-Predict Model for Face Recognition FIPA Seminar WS 2011/2012, 19.01.2012 INSTITUTE FOR ANTHROPOMATICS, FACIAL IMAGE PROCESSING AND ANALYSIS YIG University of the State of Baden-Wuerttemberg
More informationLearning to Recognize Faces in Realistic Conditions
000 001 002 003 004 005 006 007 008 009 010 011 012 013 014 015 016 017 018 019 020 021 022 023 024 025 026 027 028 029 030 031 032 033 034 035 036 037 038 039 040 041 042 043 044 045 046 047 048 049 050
More informationClass 5: Attributes and Semantic Features
Class 5: Attributes and Semantic Features Rogerio Feris, Feb 21, 2013 EECS 6890 Topics in Information Processing Spring 2013, Columbia University http://rogerioferis.com/visualrecognitionandsearch Project
More informationUnderstanding Faces. Detection, Recognition, and. Transformation of Faces 12/5/17
Understanding Faces Detection, Recognition, and 12/5/17 Transformation of Faces Lucas by Chuck Close Chuck Close, self portrait Some slides from Amin Sadeghi, Lana Lazebnik, Silvio Savarese, Fei-Fei Li
More informationAttributes and More Crowdsourcing
Attributes and More Crowdsourcing Computer Vision CS 143, Brown James Hays Many slides from Derek Hoiem Recap: Human Computation Active Learning: Let the classifier tell you where more annotation is needed.
More informationExperiments of Image Retrieval Using Weak Attributes
Columbia University Computer Science Department Technical Report # CUCS 005-12 (2012) Experiments of Image Retrieval Using Weak Attributes Felix X. Yu, Rongrong Ji, Ming-Hen Tsai, Guangnan Ye, Shih-Fu
More informationMultiple Kernel Learning for Emotion Recognition in the Wild
Multiple Kernel Learning for Emotion Recognition in the Wild Karan Sikka, Karmen Dykstra, Suchitra Sathyanarayana, Gwen Littlewort and Marian S. Bartlett Machine Perception Laboratory UCSD EmotiW Challenge,
More informationAttributes. Computer Vision. James Hays. Many slides from Derek Hoiem
Many slides from Derek Hoiem Attributes Computer Vision James Hays Recap: Human Computation Active Learning: Let the classifier tell you where more annotation is needed. Human-in-the-loop recognition:
More informationCPSC 340: Machine Learning and Data Mining. Principal Component Analysis Fall 2016
CPSC 340: Machine Learning and Data Mining Principal Component Analysis Fall 2016 A2/Midterm: Admin Grades/solutions will be posted after class. Assignment 4: Posted, due November 14. Extra office hours:
More information24 hours of Photo Sharing. installation by Erik Kessels
24 hours of Photo Sharing installation by Erik Kessels And sometimes Internet photos have useful labels Im2gps. Hays and Efros. CVPR 2008 But what if we want more? Image Categorization Training Images
More informationImageCLEF 2011
SZTAKI @ ImageCLEF 2011 Bálint Daróczy joint work with András Benczúr, Róbert Pethes Data Mining and Web Search Group Computer and Automation Research Institute Hungarian Academy of Sciences Training/test
More informationCOSC160: Detection and Classification. Jeremy Bolton, PhD Assistant Teaching Professor
COSC160: Detection and Classification Jeremy Bolton, PhD Assistant Teaching Professor Outline I. Problem I. Strategies II. Features for training III. Using spatial information? IV. Reducing dimensionality
More informationCS4495/6495 Introduction to Computer Vision. 8C-L1 Classification: Discriminative models
CS4495/6495 Introduction to Computer Vision 8C-L1 Classification: Discriminative models Remember: Supervised classification Given a collection of labeled examples, come up with a function that will predict
More informationDeep Learning for Face Recognition. Xiaogang Wang Department of Electronic Engineering, The Chinese University of Hong Kong
Deep Learning for Face Recognition Xiaogang Wang Department of Electronic Engineering, The Chinese University of Hong Kong Deep Learning Results on LFW Method Accuracy (%) # points # training images Huang
More informationObject detection using Region Proposals (RCNN) Ernest Cheung COMP Presentation
Object detection using Region Proposals (RCNN) Ernest Cheung COMP790-125 Presentation 1 2 Problem to solve Object detection Input: Image Output: Bounding box of the object 3 Object detection using CNN
More informationEquation to LaTeX. Abhinav Rastogi, Sevy Harris. I. Introduction. Segmentation.
Equation to LaTeX Abhinav Rastogi, Sevy Harris {arastogi,sharris5}@stanford.edu I. Introduction Copying equations from a pdf file to a LaTeX document can be time consuming because there is no easy way
More informationClassifying Images with Visual/Textual Cues. By Steven Kappes and Yan Cao
Classifying Images with Visual/Textual Cues By Steven Kappes and Yan Cao Motivation Image search Building large sets of classified images Robotics Background Object recognition is unsolved Deformable shaped
More informationFace Image Quality Assessment for Face Selection in Surveillance Video using Convolutional Neural Networks
Face Image Quality Assessment for Face Selection in Surveillance Video using Convolutional Neural Networks Vignesh Sankar, K. V. S. N. L. Manasa Priya, Sumohana Channappayya Indian Institute of Technology
More informationRecognition Tools: Support Vector Machines
CS 2770: Computer Vision Recognition Tools: Support Vector Machines Prof. Adriana Kovashka University of Pittsburgh January 12, 2017 Announcement TA office hours: Tuesday 4pm-6pm Wednesday 10am-12pm Matlab
More informationVignette: Reimagining the Analog Photo Album
Vignette: Reimagining the Analog Photo Album David Eng, Andrew Lim, Pavitra Rengarajan Abstract Although the smartphone has emerged as the most convenient device on which to capture photos, it lacks the
More informationLearning and Inferring Depth from Monocular Images. Jiyan Pan April 1, 2009
Learning and Inferring Depth from Monocular Images Jiyan Pan April 1, 2009 Traditional ways of inferring depth Binocular disparity Structure from motion Defocus Given a single monocular image, how to infer
More informationEnd-to-End Localization and Ranking for Relative Attributes
End-to-End Localization and Ranking for Relative Attributes Krishna Kumar Singh and Yong Jae Lee Presented by Minhao Cheng [Farhadi et al. 2009, Kumar et al. 2009, Lampert et al. 2009, [Slide: Xiao and
More informationRobust Face Recognition via Sparse Representation Authors: John Wright, Allen Y. Yang, Arvind Ganesh, S. Shankar Sastry, and Yi Ma
Robust Face Recognition via Sparse Representation Authors: John Wright, Allen Y. Yang, Arvind Ganesh, S. Shankar Sastry, and Yi Ma Presented by Hu Han Jan. 30 2014 For CSE 902 by Prof. Anil K. Jain: Selected
More informationDimensionality Reduction using Relative Attributes
Dimensionality Reduction using Relative Attributes Mohammadreza Babaee 1, Stefanos Tsoukalas 1, Maryam Babaee Gerhard Rigoll 1, and Mihai Datcu 1 Institute for Human-Machine Communication, Technische Universität
More informationTHESIS EVALUATING SOFT BIOMETRICS IN THE CONTEXT OF FACE RECOGNITION. Submitted by Hao Zhang Department of Computer Science
THESIS EVALUATING SOFT BIOMETRICS IN THE CONTEXT OF FACE RECOGNITION Submitted by Hao Zhang Department of Computer Science In partial fulfillment of the requirements For the Degree of Master of Science
More informationTracking. Hao Guan( 管皓 ) School of Computer Science Fudan University
Tracking Hao Guan( 管皓 ) School of Computer Science Fudan University 2014-09-29 Multimedia Video Audio Use your eyes Video Tracking Use your ears Audio Tracking Tracking Video Tracking Definition Given
More informationContent Based Image Retrieval
Content Based Image Retrieval R. Venkatesh Babu Outline What is CBIR Approaches Features for content based image retrieval Global Local Hybrid Similarity measure Trtaditional Image Retrieval Traditional
More informationDevelopment in Object Detection. Junyuan Lin May 4th
Development in Object Detection Junyuan Lin May 4th Line of Research [1] N. Dalal and B. Triggs. Histograms of oriented gradients for human detection, CVPR 2005. HOG Feature template [2] P. Felzenszwalb,
More informationDA Progress report 2 Multi-view facial expression. classification Nikolas Hesse
DA Progress report 2 Multi-view facial expression classification 16.12.2010 Nikolas Hesse Motivation Facial expressions (FE) play an important role in interpersonal communication FE recognition can help
More informationRecognition. Topics that we will try to cover:
Recognition Topics that we will try to cover: Indexing for fast retrieval (we still owe this one) Object classification (we did this one already) Neural Networks Object class detection Hough-voting techniques
More informationhttps://en.wikipedia.org/wiki/the_dress Recap: Viola-Jones sliding window detector Fast detection through two mechanisms Quickly eliminate unlikely windows Use features that are fast to compute Viola
More informationAnalysis: TextonBoost and Semantic Texton Forests. Daniel Munoz Februrary 9, 2009
Analysis: TextonBoost and Semantic Texton Forests Daniel Munoz 16-721 Februrary 9, 2009 Papers [shotton-eccv-06] J. Shotton, J. Winn, C. Rother, A. Criminisi, TextonBoost: Joint Appearance, Shape and Context
More informationon learned visual embedding patrick pérez Allegro Workshop Inria Rhônes-Alpes 22 July 2015
on learned visual embedding patrick pérez Allegro Workshop Inria Rhônes-Alpes 22 July 2015 Vector visual representation Fixed-size image representation High-dim (100 100,000) Generic, unsupervised: BoW,
More informationLearning Visual Semantics: Models, Massive Computation, and Innovative Applications
Learning Visual Semantics: Models, Massive Computation, and Innovative Applications Part II: Visual Features and Representations Liangliang Cao, IBM Watson Research Center Evolvement of Visual Features
More informationECG782: Multidimensional Digital Signal Processing
ECG782: Multidimensional Digital Signal Processing Object Recognition http://www.ee.unlv.edu/~b1morris/ecg782/ 2 Outline Knowledge Representation Statistical Pattern Recognition Neural Networks Boosting
More informationCPSC 340: Machine Learning and Data Mining. Multi-Dimensional Scaling Fall 2017
CPSC 340: Machine Learning and Data Mining Multi-Dimensional Scaling Fall 2017 Assignment 4: Admin 1 late day for tonight, 2 late days for Wednesday. Assignment 5: Due Monday of next week. Final: Details
More informationHistograms of Sparse Codes for Object Detection
Histograms of Sparse Codes for Object Detection Xiaofeng Ren (Amazon), Deva Ramanan (UC Irvine) Presented by Hossein Azizpour What does the paper do? (learning) a new representation local histograms of
More informationDisguised Face Identification (DFI) with Facial KeyPoints using Spatial Fusion Convolutional Network. Nathan Sun CIS601
Disguised Face Identification (DFI) with Facial KeyPoints using Spatial Fusion Convolutional Network Nathan Sun CIS601 Introduction Face ID is complicated by alterations to an individual s appearance Beard,
More informationSyllabus. 1. Visual classification Intro 2. SVM 3. Datasets and evaluation 4. Shallow / Deep architectures
Syllabus 1. Visual classification Intro 2. SVM 3. Datasets and evaluation 4. Shallow / Deep architectures Image classification How to define a category? Bicycle Paintings with women Portraits Concepts,
More informationCross-pose Facial Expression Recognition
Cross-pose Facial Expression Recognition Abstract In real world facial expression recognition (FER) applications, it is not practical for a user to enroll his/her facial expressions under different pose
More informationSURF. Lecture6: SURF and HOG. Integral Image. Feature Evaluation with Integral Image
SURF CSED441:Introduction to Computer Vision (2015S) Lecture6: SURF and HOG Bohyung Han CSE, POSTECH bhhan@postech.ac.kr Speed Up Robust Features (SURF) Simplified version of SIFT Faster computation but
More informationCS229: Action Recognition in Tennis
CS229: Action Recognition in Tennis Aman Sikka Stanford University Stanford, CA 94305 Rajbir Kataria Stanford University Stanford, CA 94305 asikka@stanford.edu rkataria@stanford.edu 1. Motivation As active
More informationCS6716 Pattern Recognition
CS6716 Pattern Recognition Aaron Bobick School of Interactive Computing Administrivia PS3 is out now, due April 8. Today chapter 12 of the Hastie book. Slides (and entertainment) from Moataz Al-Haj Three
More informationFace detection and recognition. Detection Recognition Sally
Face detection and recognition Detection Recognition Sally Face detection & recognition Viola & Jones detector Available in open CV Face recognition Eigenfaces for face recognition Metric learning identification
More informationHUMAN S FACIAL PARTS EXTRACTION TO RECOGNIZE FACIAL EXPRESSION
HUMAN S FACIAL PARTS EXTRACTION TO RECOGNIZE FACIAL EXPRESSION Dipankar Das Department of Information and Communication Engineering, University of Rajshahi, Rajshahi-6205, Bangladesh ABSTRACT Real-time
More informationCSE 158. Web Mining and Recommender Systems. Midterm recap
CSE 158 Web Mining and Recommender Systems Midterm recap Midterm on Wednesday! 5:10 pm 6:10 pm Closed book but I ll provide a similar level of basic info as in the last page of previous midterms CSE 158
More informationRich feature hierarchies for accurate object detection and semantic segmentation
Rich feature hierarchies for accurate object detection and semantic segmentation BY; ROSS GIRSHICK, JEFF DONAHUE, TREVOR DARRELL AND JITENDRA MALIK PRESENTER; MUHAMMAD OSAMA Object detection vs. classification
More informationAn Exploration of Computer Vision Techniques for Bird Species Classification
An Exploration of Computer Vision Techniques for Bird Species Classification Anne L. Alter, Karen M. Wang December 15, 2017 Abstract Bird classification, a fine-grained categorization task, is a complex
More informationRobust Facial Expression Classification Using Shape and Appearance Features
Robust Facial Expression Classification Using Shape and Appearance Features S L Happy and Aurobinda Routray Department of Electrical Engineering, Indian Institute of Technology Kharagpur, India Abstract
More informationComputer vision: models, learning and inference. Chapter 13 Image preprocessing and feature extraction
Computer vision: models, learning and inference Chapter 13 Image preprocessing and feature extraction Preprocessing The goal of pre-processing is to try to reduce unwanted variation in image due to lighting,
More informationFace/Flesh Detection and Face Recognition
Face/Flesh Detection and Face Recognition Linda Shapiro EE/CSE 576 1 What s Coming 1. Review of Bakic flesh detector 2. Fleck and Forsyth flesh detector 3. Details of Rowley face detector 4. The Viola
More informationPart-based and local feature models for generic object recognition
Part-based and local feature models for generic object recognition May 28 th, 2015 Yong Jae Lee UC Davis Announcements PS2 grades up on SmartSite PS2 stats: Mean: 80.15 Standard Dev: 22.77 Vote on piazza
More informationRecognition of Animal Skin Texture Attributes in the Wild. Amey Dharwadker (aap2174) Kai Zhang (kz2213)
Recognition of Animal Skin Texture Attributes in the Wild Amey Dharwadker (aap2174) Kai Zhang (kz2213) Motivation Patterns and textures are have an important role in object description and understanding
More informationBeyond Bags of Features
: for Recognizing Natural Scene Categories Matching and Modeling Seminar Instructed by Prof. Haim J. Wolfson School of Computer Science Tel Aviv University December 9 th, 2015
More informationThree things everyone should know to improve object retrieval. Relja Arandjelović and Andrew Zisserman (CVPR 2012)
Three things everyone should know to improve object retrieval Relja Arandjelović and Andrew Zisserman (CVPR 2012) University of Oxford 2 nd April 2012 Large scale object retrieval Find all instances of
More informationAN ENHANCED ATTRIBUTE RERANKING DESIGN FOR WEB IMAGE SEARCH
AN ENHANCED ATTRIBUTE RERANKING DESIGN FOR WEB IMAGE SEARCH Sai Tejaswi Dasari #1 and G K Kishore Babu *2 # Student,Cse, CIET, Lam,Guntur, India * Assistant Professort,Cse, CIET, Lam,Guntur, India Abstract-
More informationAnnouncements. Recognition I. Gradient Space (p,q) What is the reflectance map?
Announcements I HW 3 due 12 noon, tomorrow. HW 4 to be posted soon recognition Lecture plan recognition for next two lectures, then video and motion. Introduction to Computer Vision CSE 152 Lecture 17
More informationRobot Learning. There are generally three types of robot learning: Learning from data. Learning by demonstration. Reinforcement learning
Robot Learning 1 General Pipeline 1. Data acquisition (e.g., from 3D sensors) 2. Feature extraction and representation construction 3. Robot learning: e.g., classification (recognition) or clustering (knowledge
More informationAutomatic Colorization of Grayscale Images
Automatic Colorization of Grayscale Images Austin Sousa Rasoul Kabirzadeh Patrick Blaes Department of Electrical Engineering, Stanford University 1 Introduction ere exists a wealth of photographic images,
More informationCS229 Final Project: Predicting Expected Response Times
CS229 Final Project: Predicting Expected Email Response Times Laura Cruz-Albrecht (lcruzalb), Kevin Khieu (kkhieu) December 15, 2017 1 Introduction Each day, countless emails are sent out, yet the time
More informationComputer Vision: Summary and Discussion
12/05/2011 Computer Vision: Summary and Discussion Computer Vision CS 143, Brown James Hays Many slides from Derek Hoiem Announcements Today is last day of regular class Second quiz on Wednesday (Dec 7
More informationFACE DETECTION AND RECOGNITION OF DRAWN CHARACTERS HERMAN CHAU
FACE DETECTION AND RECOGNITION OF DRAWN CHARACTERS HERMAN CHAU 1. Introduction Face detection of human beings has garnered a lot of interest and research in recent years. There are quite a few relatively
More informationDescriptorEnsemble: An Unsupervised Approach to Image Matching and Alignment with Multiple Descriptors
DescriptorEnsemble: An Unsupervised Approach to Image Matching and Alignment with Multiple Descriptors 林彥宇副研究員 Yen-Yu Lin, Associate Research Fellow 中央研究院資訊科技創新研究中心 Research Center for IT Innovation, Academia
More informationFace detection and recognition. Many slides adapted from K. Grauman and D. Lowe
Face detection and recognition Many slides adapted from K. Grauman and D. Lowe Face detection and recognition Detection Recognition Sally History Early face recognition systems: based on features and distances
More informationMultiple-Choice Questionnaire Group C
Family name: Vision and Machine-Learning Given name: 1/28/2011 Multiple-Choice naire Group C No documents authorized. There can be several right answers to a question. Marking-scheme: 2 points if all right
More informationFeature selection. LING 572 Fei Xia
Feature selection LING 572 Fei Xia 1 Creating attribute-value table x 1 x 2 f 1 f 2 f K y Choose features: Define feature templates Instantiate the feature templates Dimensionality reduction: feature selection
More informationHeat Kernel Based Local Binary Pattern for Face Representation
JOURNAL OF LATEX CLASS FILES 1 Heat Kernel Based Local Binary Pattern for Face Representation Xi Li, Weiming Hu, Zhongfei Zhang, Hanzi Wang Abstract Face classification has recently become a very hot research
More informationHW2 due on Thursday. Face Recognition: Dimensionality Reduction. Biometrics CSE 190 Lecture 11. Perceptron Revisited: Linear Separators
HW due on Thursday Face Recognition: Dimensionality Reduction Biometrics CSE 190 Lecture 11 CSE190, Winter 010 CSE190, Winter 010 Perceptron Revisited: Linear Separators Binary classification can be viewed
More informationObject Detection with Discriminatively Trained Part Based Models
Object Detection with Discriminatively Trained Part Based Models Pedro F. Felzenszwelb, Ross B. Girshick, David McAllester and Deva Ramanan Presented by Fabricio Santolin da Silva Kaustav Basu Some slides
More informationFACE RECOGNITION BASED ON GENDER USING A MODIFIED METHOD OF 2D-LINEAR DISCRIMINANT ANALYSIS
FACE RECOGNITION BASED ON GENDER USING A MODIFIED METHOD OF 2D-LINEAR DISCRIMINANT ANALYSIS 1 Fitri Damayanti, 2 Wahyudi Setiawan, 3 Sri Herawati, 4 Aeri Rachmad 1,2,3,4 Faculty of Engineering, University
More informationAutomated Canvas Analysis for Painting Conservation. By Brendan Tobin
Automated Canvas Analysis for Painting Conservation By Brendan Tobin 1. Motivation Distinctive variations in the spacings between threads in a painting's canvas can be used to show that two sections of
More informationMORPH-II: Feature Vector Documentation
MORPH-II: Feature Vector Documentation Troy P. Kling NSF-REU Site at UNC Wilmington, Summer 2017 1 MORPH-II Subsets Four different subsets of the MORPH-II database were selected for a wide range of purposes,
More informationFace Image Retrieval Using Pose Specific Set Sparse Feature Representation
Available Online at www.ijcsmc.com International Journal of Computer Science and Mobile Computing A Monthly Journal of Computer Science and Information Technology IJCSMC, Vol. 3, Issue. 9, September 2014,
More informationRecognition: Face Recognition. Linda Shapiro EE/CSE 576
Recognition: Face Recognition Linda Shapiro EE/CSE 576 1 Face recognition: once you ve detected and cropped a face, try to recognize it Detection Recognition Sally 2 Face recognition: overview Typical
More informationObject Detection Design challenges
Object Detection Design challenges How to efficiently search for likely objects Even simple models require searching hundreds of thousands of positions and scales Feature design and scoring How should
More informationCS6501: Deep Learning for Visual Recognition. Object Detection I: RCNN, Fast-RCNN, Faster-RCNN
CS6501: Deep Learning for Visual Recognition Object Detection I: RCNN, Fast-RCNN, Faster-RCNN Today s Class Object Detection The RCNN Object Detector (2014) The Fast RCNN Object Detector (2015) The Faster
More informationSemi-supervised learning of facial attributes in video
Semi-supervised learning of facial attributes in video Neva Cherniavsky 1, Ivan Laptev 1, Josef Sivic 1, Andrew Zisserman 1,2 1 INRIA, WILLOW, Laboratoire d Informatique de l Ecole Normale Supérieure,
More informationGENDER CLASSIFICATION USING SUPPORT VECTOR MACHINES
GENDER CLASSIFICATION USING SUPPORT VECTOR MACHINES Ashwin Swaminathan ashwins@umd.edu ENEE633: Statistical and Neural Pattern Recognition Instructor : Prof. Rama Chellappa Project 2, Part (a) 1. INTRODUCTION
More informationFeature Descriptors. CS 510 Lecture #21 April 29 th, 2013
Feature Descriptors CS 510 Lecture #21 April 29 th, 2013 Programming Assignment #4 Due two weeks from today Any questions? How is it going? Where are we? We have two umbrella schemes for object recognition
More informationBilevel Sparse Coding
Adobe Research 345 Park Ave, San Jose, CA Mar 15, 2013 Outline 1 2 The learning model The learning algorithm 3 4 Sparse Modeling Many types of sensory data, e.g., images and audio, are in high-dimensional
More informationGraph based machine learning with applications to media analytics
Graph based machine learning with applications to media analytics Lei Ding, PhD 9-1-2011 with collaborators at Outline Graph based machine learning Basic structures Algorithms Examples Applications in
More informationResearch on Robust Local Feature Extraction Method for Human Detection
Waseda University Doctoral Dissertation Research on Robust Local Feature Extraction Method for Human Detection TANG, Shaopeng Graduate School of Information, Production and Systems Waseda University Feb.
More informationLecture Linear Support Vector Machines
Lecture 8 In this lecture we return to the task of classification. As seen earlier, examples include spam filters, letter recognition, or text classification. In this lecture we introduce a popular method
More informationCPSC 340: Machine Learning and Data Mining. Deep Learning Fall 2018
CPSC 340: Machine Learning and Data Mining Deep Learning Fall 2018 Last Time: Multi-Dimensional Scaling Multi-dimensional scaling (MDS): Non-parametric visualization: directly optimize the z i locations.
More informationLIG and LIRIS at TRECVID 2008: High Level Feature Extraction and Collaborative Annotation
LIG and LIRIS at TRECVID 2008: High Level Feature Extraction and Collaborative Annotation Stéphane Ayache, Georges Quénot To cite this version: Stéphane Ayache, Georges Quénot. LIG and LIRIS at TRECVID
More informationThe CSU Face Identification Evaluation System: Its Purpose, Features and Structure
The CSU Face Identification Evaluation System: Its Purpose, Features and Structure David S. Bolme, J. Ross Beveridge, Marcio Teixeira and Bruce A. Draper Computer Science, Colorado State University 3rd
More informationLecture 12 Recognition
Institute of Informatics Institute of Neuroinformatics Lecture 12 Recognition Davide Scaramuzza 1 Lab exercise today replaced by Deep Learning Tutorial Room ETH HG E 1.1 from 13:15 to 15:00 Optional lab
More informationPreviously. Window-based models for generic object detection 4/11/2011
Previously for generic object detection Monday, April 11 UT-Austin Instance recognition Local features: detection and description Local feature matching, scalable indexing Spatial verification Intro to
More informationDeep Learning for Computer Vision
Deep Learning for Computer Vision Spring 2018 http://vllab.ee.ntu.edu.tw/dlcv.html (primary) https://ceiba.ntu.edu.tw/1062dlcv (grade, etc.) FB: DLCV Spring 2018 Yu Chiang Frank Wang 王鈺強, Associate Professor
More informationLarge-Scale Traffic Sign Recognition based on Local Features and Color Segmentation
Large-Scale Traffic Sign Recognition based on Local Features and Color Segmentation M. Blauth, E. Kraft, F. Hirschenberger, M. Böhm Fraunhofer Institute for Industrial Mathematics, Fraunhofer-Platz 1,
More informationTri-modal Human Body Segmentation
Tri-modal Human Body Segmentation Master of Science Thesis Cristina Palmero Cantariño Advisor: Sergio Escalera Guerrero February 6, 2014 Outline 1 Introduction 2 Tri-modal dataset 3 Proposed baseline 4
More informationApplications Video Surveillance (On-line or off-line)
Face Face Recognition: Dimensionality Reduction Biometrics CSE 190-a Lecture 12 CSE190a Fall 06 CSE190a Fall 06 Face Recognition Face is the most common biometric used by humans Applications range from
More informationHybrid Deep Learning for Face Verification. Yi Sun, Xiaogang Wang, Member, IEEE, and Xiaoou Tang, Fellow, IEEE
1 Hybrid Deep Learning for Face Verification Yi Sun, Xiaogang Wang, Member, IEEE, and Xiaoou Tang, Fellow, IEEE Abstract This paper proposes a hybrid convolutional network (ConvNet)-Restricted Boltzmann
More informationVEHICLE CLASSIFICATION And License Plate Recognition
VEHICLE CLASSIFICATION And License Plate Recognition CS771A Course Project : Under Prof. Harish Karnick Amlan Kar Nishant Rai Sandipan Mandal Sourav Anand Group 26 Indian Institute of Technology Kanpur
More informationCPSC 340: Machine Learning and Data Mining. Principal Component Analysis Fall 2017
CPSC 340: Machine Learning and Data Mining Principal Component Analysis Fall 2017 Assignment 3: 2 late days to hand in tonight. Admin Assignment 4: Due Friday of next week. Last Time: MAP Estimation MAP
More informationScalable Face Image Retrieval Using Attribute-Enhanced Sparse Codewords
IEEE TRANSACTIONS ON MULTIMEDIA, VOL. 15, NO. 5, AUGUST 2013 1163 Scalable Face Image Retrieval Using Attribute-Enhanced Sparse Codewords Bor-Chun Chen, Yan-Ying Chen, Yin-Hsi Kuo, and Winston H. Hsu,
More information